Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 31 Jan 2003 13:09:39 +0100
From:      Alexander Leidinger <Alexander@Leidinger.net>
To:        Matthew Dillon <dillon@apollo.backplane.com>
Cc:        jhb@FreeBSD.org, cvs-all@FreeBSD.org, cvs-committers@FreeBSD.org
Subject:   Re: cvs commit: src/sys/i386/i386 mp_machdep.c
Message-ID:  <20030131130939.460cef5e.Alexander@Leidinger.net>
In-Reply-To: <200301301952.h0UJqA0h089352@apollo.backplane.com>
References:  <XFMail.20030130114117.jhb@FreeBSD.org> <200301301952.h0UJqA0h089352@apollo.backplane.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 30 Jan 2003 11:52:10 -0800 (PST)
Matthew Dillon <dillon@apollo.backplane.com> wrote:

>     Also, I perused the linux archives a bit and they had some rather
>     severe scheduling issues with HTT.  Apparently the logical cpus are
>     not equivalent to each other (?), and are definitely not as effective
>     as physical cpus (which makes sense), so if you have two KSEs and two
>     physical cpus each with an extra logical cpu, you absoultely want to
>     schedule the two KSEs on the two physical cpu's.

The logical CPUs share the same execution units, they just have their
own set of registers (this is a simplified view). Ideally you should
balance KSEs with different unit usage characteristics on the same
physical CPU (which is not possible to determine) because two integer
intensive KSEs on the same physical CPU means that one KSE has to wait
until the other KSE has finished his actual instruction (it doesn't
matter if the instructions get mixed like the teeth of a zipper or not,
the end-result is like scheduling the KSEs on a UP system), whereas a
integer and a floating-point intensive KSE can get a performance boost
when run on two logical CPUs which reside on one physical CPU.

Bye,
Alexander.

-- 
            Secret hacker rule #11: hackers read manuals.

http://www.Leidinger.net                       Alexander @ Leidinger.net
  GPG fingerprint = C518 BC70 E67F 143F BE91  3365 79E2 9C60 B006 3FE7

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030131130939.460cef5e.Alexander>